Frames of depth maps of a scene can be generated by processing depth measurements acquired by a depth sensing device. However, the frames of depth maps are usually generated at a much slower rate than are image frames of a video of that same scene acquired using a video camera. It is sometimes desired to process depth frames generated of a scene so they have a same or similar frame rate (temporal resolution) as that of the video of that scene. Moreover, the spatial resolution of the generated depth maps tends to be smaller than the spatial resolution of the video images of that scene. Therefore, it is desirable to be able to spatially and temporally enhance a sequence of depth frames of a scene.
Accordingly, what is disclosed is a system and method for enhancing a spatio-temporal resolution of time-sequential depth frames of a scene.